STRUCTURE AND FUNCTION OF TRACHEARY ELEMENTS IN AMBORELLA TRICHOPODA

Transcription

1 Int. J. Plant Sci. 161(5): by The University of Chicago. All rights reserved /2000/ $03.00 STRUCTURE AND FUNCTION OF TRACHEARY ELEMENTS IN AMBORELLA TRICHOPODA Taylor S. Feild, 1, * Maciej A. Zweiniecki,* Tim Brodribb, Tanguy Jaffré, Michael J. Donoghue,* and N. Michele Holbrook* *Department of Organismic and Evolutionary Biology, Harvard University, Cambridge, Massachusetts 02138, U.S.A.; Department of Plant Science, University of Tasmania, GPO Box , Hobart 70001, Australia; and Laboratoire de Botanique, Institut de Recherche pour le Développement., B.P. A5, Noumeá Cedex, New Caledonia Recent phylogenetic analyses have placed the root of flowering plants near Amborella trichopoda, a woody plant restricted to cloud forest habitats in New Caledonia. A distinctive feature of A. trichopoda is its reported lack of xylem vessels. Here we present observations of pit membrane structure and end wall morphology for primary and secondary tracheary cells of A. trichopoda as well as field measurements of stem hydraulic properties of A. trichopoda compared with five cloud fforest species from New Caledonia. Observations of stem radial sections revealed that the primary wall material in the protoxylem and metaxylem elements was intact. No large porosities (such as those that have been observed in the pit membranes of Nymphaeales) were observed. However, a few elliptical pits of tracheary cells in the secondary xylem appeared to lack pit membranes. These observations are consistent with our measurements of functional conduit length, which indicate that the longest open conduits are equal to the length of two secondary xylem elements joined end to end. Thus, the xylem of A. trichopoda appears to be functionally vesselless, with the caveat that connections between individual vascular elements may occasionally be open (i.e., lacking in at least one pit membrane). Sapwood area and leaf area specific hydraulic conductivities of A. trichopoda are similar to those of conifers and angiosperms, with and without xylem vessels, growing in understory cloud forest environments. These findings bear on discussions of the morphology and ecology of the first flowering plants as well as on the possible causes of their diversification. Keywords: Amborella, basal angiosperms, xylem hydraulic conductivity, vessel evolution. Introduction Botanists have long been fascinated by the origin and radiation of flowering plants (Takhtajan 1969; Thorne 1974; Doyle 1978; Doyle and Donoghue 1986; Cronquist 1988; Donoghue and Doyle 1989; Taylor and Hickey 1992; Crane et al. 1995; Wing and Boucher 1998). Recent phylogenetic analyses of multiple molecular data sets have resolved the root of flowering plants near Amborella, water lilies (i.e., Nymphaeales), and a clade containing Austrobaileya, Trimeniaceae, and Illiciaceae plus Schisandraceae (Mathews and Donoghue 1999; Parkinson et al. 1999; Qiu et al. 1999; Soltis et al. 1999). These results bear directly on the evolutionary direction of morphological change (e.g., in carpel morphology, habit, breeding system, tracheary elements) and on ecological shifts that may have contributed to the rapid diversification of angiosperms in the Early to Middle Cretaceous (Doyle and Donoghue 1986; Crane and Ligard 1989; Taylor and Hickey 1992; Doyle and Donoghue 1993; Sanderson and Donoghue 1994; Crane et al. 1995; Doyle 1998; Wing and Boucher 1998; Mathews and Donoghue 1999). One important implication of placing Amborella as the sister group of all other angiosperms 1 Author for correspondence. Telephone ; fax ; tfeild@oeb.harvard.edu. Manuscript received January 2000; revised manuscript received April is that it increases the likelihood that the first flowering plants were woody plants with vesselless xylem (i.e., containing only tracheids for water transport; contra Young 1981; Donoghue and Doyle 1989; Baas and Wheeler 1996; see also Feild et al. 1998). Amborella trichopoda Baill., the only species of Amborellaceae, is an understory to subcanopy shrub (maximally 8 m tall) growing in middle-elevation ( m above sea level) montane tropical rain forests of central New Caledonia, Australia (Jérémie 1980). Although it is reported to lack vessels (Bailey and Swamy 1948; Bailey 1957; Metcalfe 1987), the discovery of incipient vessel elements in some water lilies (Schneider and Carlquist 1995a, 1995b) and in other reputed vesselless angiosperms (e.g., Sarcandra, Chloranthaceae; Carlquist 1987; Takahashi 1988) underscores the need for a careful reexamination of A. trichopoda pit membrane structure and the end wall morphology of tracheary cells. Currently, no information is available on the water-transport capacity or relevant ecological traits of A. trichopoda in relation to woody plants living in the same communities. In this article, we provide the first observations of pit membrane structure for tracheary cells of A. trichopoda using scanning electron microscopy (SEM). We also report field measurements on the water-transport capacity of A. trichopoda stems compared to those of vessel-bearing trees and vesselless conifers and vesselless angiosperms (represented by Winteraceae) in New Caledonia. Our findings bear on discussions of the morphology 705

2 706 INTERNATIONAL JOURNAL OF PLANT SCIENCES and ecology of the first flowering plants and on the possible causes of their diversification. Material and Methods Study Areas and Species Examined For anatomical and physiological studies, we sampled seven shrubs (each 1 2 m tall) of Amborella trichopoda from a large population (ca. 500 plants) growing from 600 to 800 m on Plateau de Dogny (22 35 S; E) near Sarameá, New Caledonia, on July 27, Leaves on the measured branches were free of epiphylls, thereby ensuring that these leaves would be operating at maximum hydraulic capacity. The climate at this site is tropical, characterized by a wet season that extends from December to March, a short dry season that extends from April to May, a wet season that extends from June to August, and another dry season that extends from September to November (Jaffré 1980). Annual rainfall and temperature variation on Plateau de Dogny have not been measured; however, rainfall totals in Sarameá may approach 2500 mm a year (METEO-France 1994). The forest is often enshrouded in mist, and cloud moisture contributes a significant but unknown amount to overall moisture input (T. S. Feild, personal observation). Temperatures in New Caledonia are mild year-round; the mean monthly temperature in Noumeá (140 km away) is highest in February (26.2 C) and lowest in August (19.9 C; Jaffré 1980). To put the hydraulic properties of A. trichopoda in context, we also measured the stem hydraulic properties of two conifers (Agathis lanceolata Warburg. [Araucariaceae] and Retrophyllum comptonii C. N. Page [Podocarpaceae]), two members of the vesselless angiosperm family Winteraceae (Belliolum crassifolium [Baillon] van Tieghem and Zygogynum bailonii van Tieghem), and one vessel-forming angiosperm (Nothofagus condonandra [Baill.] van Steenis [Nothofagaceae]) growing in an understory environment of a similar tropical montane rain forest on Mt. Dzumac (22 3 S; E; 980 m), Province Sud, New Caledonia; measurements took place from late July to early August Mt. Dzumac was chosen because it was possible to establish a laboratory for measuring stem hydraulics, thus avoiding the transport of stems to Noumeá for measurement purposes. Climate data are limited for this site, with annual rainfall totaling 1816 mm in 1996 and 2742 mm in 1997; ca. one-half of this rainfall occurred from December through March (Enright and Goldblum 1998). Xylem Anatomical Observations and Functional Conduit Length Observations on A. trichopoda tracheary element dimensions (diameter and length) were made on stems preserved in 50% ethyl alcohol. Observations were made within 3 mo of preservation. Before sectioning, stem portions were rehydrated (in a stepwise manner) from 50% to 30% ethyl alcohol for 1 h and then to distilled water for 1 h. Tracheary element lengths were measured on macerated samples of stem xylem using standard procedures (O Brien and McCully 1981). Average tracheary cell diameter and length were determined from 100 measurements on macerated cells from three stems of similar diameter (17 20 mm) from three plants. For observations of pit membrane ultrastructure, transverse, tangential, and radial sections were cut from A. trichopoda stems. These sections were then mounted on aluminum stubs, air-dried, sputtercoated with platinum, and examined with a digital scanning electron microscope (LEO 495) at 5 kv. The length of the longest functional conduits was determined by forcing air at low pressure (ca MPa) through stem segments (Zimmermann and Jeje 1981). Fresh stems were fitted to a hand pump and slowly pressurized to 0.01 MPa. The stem segments were repeatedly shortened by 5 mm (using a razor blade) until the first bubbles appeared through the downstream end. This method is based on the assumptions that extracellular passageways permeable to air are not present within the stem and that gas cannot pass through wet pit membranes except at relatively high (typically greater than 1.5 MPa) pressures. In contrast, a conduit (tracheid or vessel) cut open at both ends can conduct gas at low (less than 0.01 MPa) pressure (Zimmermann and Jeje 1981). Stem Hydraulic Measurements We used a low-pressure steady state flowmeter (SSFM) method to measure stem hydraulic conductivity (K H ; Tyree et al. 1993; Zwieniecki et al. 2000; Feild and Holbrook, in press). This method, which uses the drop in pressure across a tube of known conductance to measure the flow rate into the stem segment, was chosen over the traditional methods for measuring stem K H (in which an analytical balance is used) because of the remoteness of the field sites. Briefly, an elevated reservoir (30 60 cm) was used to drive water through a 1.25-m-long piece of capillary tubing (PEEK, Upchurch Scientific, Seattle) and the stem segment, arranged in a series. The reservoir water was filtered through a 0.01-mm filter before reaching the stem. Two pressure transducers (PX236, Omega Engineering, Stamford, Conn.) were used to measure the pressure at each end of the capillary tube, allowing us to determine both the pressure drop across the capillary tube and the delivery pressure to the stem segment. The capillary tube could be easily interchanged with PEEK capillary tubes of different inner diameters (and lengths) so that the conductance of the tube was approximately equal to that of the stem. The conductance of each tube was determined in the laboratory by comparing the pressure drop across the tube at a known temperature to the outflow rate measured by an analytic balance ( 0.01 mg). Five delivery pressures (0.001, , , 0.007, and MPa) were used, and a linear calibration equation was determined for each capillary tube. Capillary tubes were calibrated before travel to New Caledonia and then immediately upon return to check for drift. After correcting for temperature, a 2% change was observed in the calibration value over a 2- mo period. Stems were measured under delivery pressures ranging from to MPa. In this range, mass flux is linearly related to applied pressure for stems (Sperry et al. 1988; Chiu and Ewers 1993). Data were recorded when steady state pressure readings were achieved, generally ca. 10 min after connecting the stem segment to the flowmeter. During all measurements, stem temperature was measured with a copperconstantan thermocouple (30 American wire gauge, Omega Engineering) so that the effect of temperature on viscosity of

3 water flux could be removed by normalizing all measurements to 25 C (Sperry et al. 1988). As a result of logistical constraints, we were unable to provide a high-pressure flush to establish maximum stem hydraulic conductivity (Sperry et al. 1988). Instead, branches were sampled at times when losses of hydraulic conductivity attributable to embolisms were likely to be at a minimum (i.e., early morning [ hours] on wet, cloudy days). For A. trichopoda, branches were cut in the field, triple-bagged in plastic, and transported immediately to a laboratory in Noumeá. On Mt. Dzumac, we were able to sample branches and immediately process them in a laboratory that had been set up in a tent. Stem segments that were longer than the longest functional conduit were cut from branches underwater (Sperry et al. 1988). Maximum conduit length for all species had been previously determined, as described above, from a minimum of three branches for each species. Both ends of the stem segment were recut with a fresh razor blade before attachment to the SSFM. Sapwood area specific conductivity (K S ; kg MPa 1 m 1 s 1 ) was determined by dividing stem K H by the sapwood cross-section area. Sapwood cross-section was measured with a micrometer along the major and minor axes of the distal stem end, and the surface area contributed by the pith was subtracted (Sperry et al. 1988). Measured branches were!2 cm in diameter, and all of the wood was considered to be sapwood (i.e., functional for water transport). Leaf specific conductivity (K L ;kgmpa 1 m 1 s 1 ) was determined by dividing K H by the total leaf area supported distal to the measured segment. For leaf area measurements, individual leaves of each branch were traced (flat) onto paper. These leaf silhouettes were subsequently measured with a leaf area meter (Li-1000, Li-Cor, Lincoln, Nebr.). Huber value was expressed as the ratio of sapwood area (in m 2 ) to leaf area (in m 2 ). Results FEILD ET AL. AMBORELLA TRACHEARY ELEMENTS 707 The primary xylem of Amborella trichopoda consisted of a three to seven cell-thick band around the large pith of the stem (figs. 1, 2). Primary xylem conduits were thin walled ( mean p 1.2 mm; standard deviation [SD] p 0.2 mm) and varied considerably in diameter from 29 (protoxylem elements) to 69 mm (metaxylem cells). The mean diameter of primary xylem conduits was 48 mm ( SD p 11.6 mm). Consistent with earlier reports, we also found numerous large (ca. 100-mm) cavities near the primary xylem band in A. trichopoda stems (figs. 1, 2; see Bailey 1957). Secondary tracheary cells occurred in relatively uniform files (in terms of lumen diameter and wall thickness) radiating from the center of the stem (fig. 1). The diameter of secondary xylem tracheary elements of A. trichopoda stems ranged from 20 to 49 mm ( mean p 28 mm; SD p 5.8 mm). Sec- ondary walls were thicker than primary xylem walls by a measure of as much as 5 mm ( mean p 4.2 mm; SD p 0.2 mm). Lengths of secondary xylem tracheary elements ranged from 1049 to 2400 mm ( mean p 1524 mm; SD p 210 mm). We observed substantial variation in the morphology, frequency, and position of tracheary element pitting in both the primary and secondary xylem. In the primary xylem, the latestformed metaxylem elements had numerous variably sized and shaped pits (fig. 2A, 2B). Metaxylem pits were elongate slits and occurred in one or more rows along the length of the cells Fig. 1 Stem cross sections of Amborella trichopoda showing the organization of primary and secondary tracheary elements. A, Overall vascular pattern of A. trichopoda, showing relatively homogeneous cell diameters and thicker walls of the secondary tracheary cells as compared to the thinner-walled primary xylem strands located near the pith. B, Detail of the primary to secondary xylem transition and structure of the relatively large lacuna near the pith. Scale bar p 150 mm ina and 30 mm inb. (fig. 2B). In the secondary xylem, pit shape ranged from elliptical to scalariform (fig. 2C), and these pits were also arranged in one or more rows along the radial walls. It was not uncommon to find both elliptical and scalariform pits clustered at the ends of secondary xylem tracheary elements (fig. 2D). Radial sections revealed that scalariform pitting occurred sporadically in A. trichopoda stems; there was no tendency for scalariform pitting to be expressed in younger versus older xylem. SEM observations of A. trichopoda xylem structure were generally consistent with a tracheid-based conducting system. Primary wall material between the gyres of the thickening bands in the primary xylem did not contain any visible perforations. Except for what appeared to be damage during sample preparation (i.e., obvious tears or cuts), the pit membranes

4 708 INTERNATIONAL JOURNAL OF PLANT SCIENCES Fig. 2 SEM photographs of the protoxylem to metaxylem transition and tracheary element pitting morphology of primary and secondary xylem of Amborella trichopoda stems from radial sections. A, Detail of the transition from protoxylem elements (helical rings) to metaxylem cells in the primary xylem. The primary wall material between the helical bands of protoxylem cells was torn away during sectioning. Note the abundance of pits on metaxylem elements located in the center. B, Latest-formed metaxylem tracheary cells with numerous slitlike pits containing intact pit membranes. C, Detail of intact primary wall material (pit membrane) of metaxylem tracheary cells; irregularly shaped holes are likely to be tears in the pit membrane resulting from sectioning. D, Co-occurrence of elliptical and scalariform pits on the end wall portions of secondary tracheary cells. Scale bar p 15 mm ina and B and 10 mm inc and D. of all primary and most secondary xylem elements were smooth and intact (fig. 3A 3C). A few elliptical and scalariform pits exhibited visible pores, although none of these exceeded 0.2 mm in diameter (fig. 3A). In a few instances, groups of bordered pits appeared to lack pit membranes altogether (fig. 3D). Although the potential for artifacts arising from sample preparation can never be fully eliminated, we believe that the disappearance of pit membranes (a disappearance not accompanied by any visible remnant) is unlikely to have been caused by either air-drying or sectioning. Consistent with the observation that intertracheid membranes were largely intact, low-pressure air (0.01 MPa) bubbled through the xylem of A. trichopoda when stem sections were shortened to a length that was less than the summed average length of two secondary xylem elements joined end to end (i.e., ca. 5 mm). This indicates that functional conduits in A. trichopoda are, at most, composed of two xylem elements. In addition, because open pits (i.e., bordered pits that lack a pit membrane) were not frequently observed, the typical conduit is likely to consist of a single tracheary cell. We observed threefold variation in stem hydraulic properties across the conifers, vesselless angiosperms, and vessel-bearing angiosperms sampled from the forest understory (table 1). The capacity of the xylem system to conduct water (K S, sapwood normalized hydraulic conductivity) varied from 0.27 kg MPa 1 m 1 s 1 in Agathis lanceolata to 0.86 kg MPa 1 m 1 s 1 in Nothofagus codonandra. Within this range, the K S of A. trichopoda stems was ca. 70% of the maximum K S measured in N. codonandra (table 1). The hydraulic capacity of A. trichopoda wood was greater than that of both species of conifers and of the Winteraceae that we sampled (table 1). The ability of A. trichopoda stems to supply a supported leaf area with water (leaf-specific hydraulic conductivity, K L ) was ca. 60% of the K L measured in vessel-forming N. condonandra (table 1). Leaf specific hydraulic conductivity is determined by the xylem hydraulic capacity (K S ) as well as by the relative in-

5 FEILD ET AL. AMBORELLA TRACHEARY ELEMENTS 709 Fig. 3 SEM photographs of variation in pit membrane intactness of Amborella trichopoda stems observed from radial and cross-sectional views. A, Presence of intact pit membranes on elliptical pits located on secondary xylem tracheary elements. B, Cross section of an intact intertracheary elliptical pit of the secondary xylem. C, Detail of small porosities in the pit membranes of secondary tracheary cells. D, Pit membranes completely lacking within intertracheary elliptical pits in the secondary xylem. Scale bars p 8 mm ina and B, 2.5 mm inc, and 16 mm ind. vestment of conducting xylem area (sapwood) to leaf area (i.e., Huber value; Zimmermann 1983; Tyree and Ewers 1991, 1996). We observed a relatively large variation in the ratio of sapwood to leaf area allocation among species; this ratio ranged from 2.0 in A. lanceolata to 3.3 in Nothofagus codonandra The Huber values of A. trichopoda were well within the range of all plants measured (table 1). Discussion The presence or absence of xylem vessels is, in general, rather easily determined, because the characteristic features of vessels (absence of pit membranes, short length and wide diameter of vessel elements, interconnection of a large number of vessel elements, differentiated perforation plates, etc.) tend to be tightly correlated (Bailey and Tupper 1918; Frost 1930; Herendeen et al. 1999). Likewise, in most cases the functional consequences of these structural differences are quite obvious. In general, plants with vessels have greater water-transport efficiency because vessels are multicelled water-conducting compartments, with the axial cell walls partially or wholly removed during development (Frost 1930; Carlquist 1975; Zimmermann and Jeje 1981; Zimmermann 1983). In contrast, tracheids are smaller in diameter and are single cells bounded by intact pit membranes (Zimmermann 1983). Pit membranes, although porous, can contribute a substantial resistance to water flow (Carlquist 1975; Zimmermann 1983; Bond 1989; Tyree and Ewers 1991, 1996). Consequently, vessel-forming plants have greater hydraulic conductance, because water passes through larger-diameter tubes and because fewer pit membranes are traversed during transport (Zimmermann 1983; Tyree and Ewers 1991, 1996). In addition, the evolutionary appearance of xylem vessels is accompanied by the development of vascular elements specialized for support (i.e., fibers) and an increased abundance of xylem parenchyma. The greater hydraulic efficiency of xylem vessels may allow for improvements in nonhydraulic aspects of stem function (e.g., support, storage) without a loss in overall conducting capacity. Our observations of Amborella trichopoda, combined with studies of other species representing early diverging angio-

6 710 INTERNATIONAL JOURNAL OF PLANT SCIENCES Table 1 Stem Hydraulic Properties of Amborella trichopoda Compared to Tropical Conifers, Vesselless Angiosperms, and a Vessel-Forming Angiosperm Growing in the Understory of New Caledonian Tropical Montane Cloud Forests Species Family K L # 10 4 K S HV # 10 4 Conifers: Agathis lanceolata (n p 4)... Araucariaceae Retrophyllum comptonii (n p 4)... Podocarpaceae Vesselless angiosperms: Amborella trichopoda (n p 10)... Amborellaceae Belliolum crassifolium (n p 10)... Winteraceae Zygogynum baillonii (n p 10)... Winteraceae Vessel-forming angiosperms: Nothofagus codonandra (n p 6)... Nothofagaceae Note. Averages of leaf area specific hydraulic conductivity (K L ;kgmpa 1 m 1 s 1 [ SD]), stem conductance per unit sapwood area (K S ; kg MPa 1 m 1 s 1 [ SD]), and Huber value (HV; ratio of sapwood area [m 2 ] to leaf area [m 2 ] [SD]), with sample sizes (n) indicated for each species. sperm lineages (Carlquist 1987, 1990, 1992, 1996; Takahashi 1988; Schneider and Carlquist 1995a, 1995b), indicate that the set of characteristics that distinguishes typical tracheids from typical vessels may not be so tightly linked in these plants. Various combinations of xylem features that in effect delimit a spectrum of structural and functional conditions that are intermediate between tracheids, as exemplified by conifers and the vessel elements typical of most angiosperms, have been documented. For example, many water lilies (Schneider and Carlquist 1995a, 1995b) and the monocot Acorus (Carlquist and Schneider 1997) exhibit large pores in their pit membranes, whereas in the case of Illicium and most Chloranthaceae, remnants of pit membranes are present in scalariform perforation plates (Carlquist 1987, 1990, 1992). The functional consequences (e.g., for embolisms and hydraulic flux) of these structural differences have not yet been studied but may result in a range of intermediate hydraulic capacities among basal lines of angiosperms. Under these circumstances, rather than trying to categorize the first angiosperms as either having or lacking vessels (e.g., Bailey and Tupper 1918; Young 1981; Carlquist 1983, 1987; Donoghue 1989; Baas and Wheeler 1996), it may be more productive to simply describe the morphological characteristics and functional attributes observed in representatives of the early angiosperm lineages. Such a decomposition of states may allow the best inference of the structure and function of tracheary elements in ancestral flowering plants. From this perspective, most characteristics of A. trichopoda match a stereotyped tracheid-bearing system. In contrast to secondary xylem in a typical vessel-bearing plant, there is little differentiation in cell diameter or wall thickness of tracheary elements. Likewise, perforation plates are not clearly demarcated, although secondary tracheary cells do show a tendency toward differentiation of end wall pitting. The majority of pit membranes are structurally and functionally intact. The porosities observed in a few pit membranes are similar in size to those reported in some Winteraceae (Carlquist 1983), but they are more than an order of magnitude smaller than those found in Nymphaeales and Acorus (Schneider and Carlquist 1995a, 1995b; Carlquist and Schneider 1997). Although bordered pits lacking pit membranes were occasionally observed, measurements of functional conduit length indicate that at most, two tracheary cells are directly connected. Therefore, with respect to the spread of embolism, the A. trichopoda system may function more like a system that contains only tracheids. However, our observation that pit membranes are sometimes absent indicates that the capacity to hydrolyze pit membranes is present in secondary xylem of A. trichopoda. Amborella trichopoda appears to be intermediate in terms of water transport. We have shown that the maximum xylem hydraulic capacity of A. trichopoda exceeds that of all other tracheid-containing plants growing in the understory of the tropical premontane rain forests sampled. In fact, A. trichopoda falls in the upper range of capacities reported previously for tropical and temperate conifers and Winteraceae (table 1; Tyree and Ewers 1991, 1996; Becker et al. 1999; Brodribb and Hill 1999; Feild and Holbrook, in press). However, most vessel-forming plants have stem conductivities that are at least two times greater than that of A. trichopoda (Tyree and Ewers 1991, 1996). To understand the significance of these findings, it is important to appreciate the environmental context. Amborella trichopoda is a shrub that is adapted to the wet understory of rain forests in New Caledonia (Bailey and Swamy 1948; Jérémie 1980). The absence of growth rings and the high abundance of epiphylls on older leaves indicate that A. trichopoda grows in uniformly wet conditions throughout the year (T. S. Feild, personal observations). In addition, A. trichopoda seedlings are most abundant in the understory and not in patches of rain forest that have been disturbed by tree falls, cyclone activity, and/or fire (T. S. Feild, personal observation). What do these observations indicate about the ecological circumstances surrounding the origin of angiosperms and of vessels? The phylogenetic position and current ecology of A. trichopoda seem inconsistent with hypotheses that the first angiosperms were gap-colonizing weed trees or rhizomatous, herbaceous plants living in seasonally dry riparian or disturbed habitats (e.g., Stebbins 1974; Hickey and Doyle 1977; Doyle 1978; Taylor and Hickey 1992; Doyle and Donoghue 1993). We note, however, that the morphology and ecology of A. trichopoda may not have been retained from the ancestor of angiosperms; instead, they might have originated more recently within the Amborella lineage. A long period of time (probably more than 140 million yr) separates the origin of the Amborella branch from A. trichopoda as we know it

7 FEILD ET AL. AMBORELLA TRACHEARY ELEMENTS 711 today. Furthermore, the rain forests in which A. trichopoda now grows in New Caledonia are not likely to have been climatically and geologically stable refugia for the Amborella line since its origin. Indeed, New Caledonia may have been entirely submerged (or nearly so) during the Paleocene to Eocene, and likewise, the emplacement of ultramafic substrates during the Eocene dramatically altered the landscape of New Caledonia (Brothers and Lillie 1988; Jaffré 1995). These observations imply that many of the so-called primitive angiosperms (e.g., Winteraceae, Trimenia, etc.) now present in New Caledonia must have arrived there more recently (Pole 1994). The same may be true of presumed relicts elsewhere, such as Austrobaileya (Endress 1983). In general, we caution against interpreting A. trichopoda as a model of the first angiosperm. Finally, we note that the inference of ancestral conditions in angiosperms must take into account the lineage that includes all angiosperms except Amborella, within which the water lily lineage and the Austrobaileya-Trimeniaceae-Illiciales lineage appear to have diverged rather early. The water lily lineage may represent an early shift into freshwater aquatic habitats accompanied by loss of secondary growth (and possibly loss of typical vessels). These plants may now be specialized to the point that their vascular and ecological traits have little bearing on conditions in the first angiosperms. The Austrobaileya-Trimeniaceae-Illiciales clade is noteworthy for the widespread occurrence of woody vines (Smith 1947; Bailey and Swamy 1949; Carlquist 1984). This occurrence may reflect a primary radiation as lianas, perhaps accompanied by the acquisition (or stabilization) and further elaboration of a suite of vessel characters associated with the transport of water through slender stems. Vines generally have a high intrinsic capacity for water transport because of the presence of large-diameter vessels and long vessels (Ewers et al. 1991; Gartner 1991). Evolution of the climbing habit in this lineage may have been foreshadowed to some extent by A. trichopoda, which has a scandent habit (as in Austrobaileya), marked by the production of leaning branches that come close to touching the ground, and reflexed petioles (Jérémie 1980). Acknowledgments We thank H. S. Feild for field assistance in New Caledonia. We are also grateful to J. A. Doyle for helpful discussions and to M. V. Thompson, A. R. Cobb, and P. J. Melcher for constructive comments on the manuscript. This research was supported by a dissertation improvement grant to Taylor S. Feild (IBN ) and by the Andrew W. Mellon Foundation. Literature Cited Baas P, EA Wheeler 1996 Parallelism and reversibility in xylem evolution a review. IAWA Bull 17: Bailey IW 1957 Additional notes on the vesselless dicotyledon, Amborella trichopoda. J Arnold Arbor Harv Univ 35: Bailey IW, BGL Swamy 1948 Amborella trichopoda Baill., a new morphological type of vesselless dicotyledon. J Arnold Arbor Harv Univ 29: The morphology and relationships of Austrobaileya. J Arnold Arbor Harv Univ 30: Bailey IW, WW Tupper 1918 Size variation in tracheary cells. I. A comparison between the secondary xylems of vascular cryptogams, gymnosperms and angiosperms. Proc Am Acad Arts 54: Becker PM, MT Tyree, M Tsuda 1999 Hydraulic conductances of angiosperms versus conifers: similar transport sufficiency at the whole plant level. Tree Physiol 19: Bond WJ 1989 The tortoise and the hare: ecology of angiosperm dominance and gymnosperm persistence. Biol J Linn Soc 36: Brodribb T, RS Hill 1999 The importance of xylem constraints in the distribution of conifer species. New Phytol 143: Brothers RN, AR Lillie 1988 Regional geology of New Caledonia. Pages in AEM Nairin, GF Stehli, S Uyeda, eds. The ocean basin and margins. Vol 7B. Plenum, New York. Carlquist S 1975 Ecological strategies of xylem evolution. University of California Press, Berkeley. 259 pp Wood anatomy of Bubbia (Winteraceae); with comments on origin of vessels in dicotyledons. Am J Bot 70: Wood anatomy of Trimeniaceae. Plant Syst Evol 144: Presence of vessels in wood of Sarcandra (Chloranthaceae); comments on the vessel origin in angiosperms. Am J Bot 74: Wood anatomy of Ascarina (Chloranthaceae) and the tracheid-vessel element transition. Aliso 12: Pit membrane remnants in perforation plates of primitive dicotyledons and their significance. Am J Bot 79: Carlquist S, EL Schneider 1997 Origins and nature of vessels in monocotyledons. I. Acorus. Int J Plant Sci 158: Chiu ST, FW Ewers 1993 The effect of segment length on conductance measurements in Lonicera fragrantissima. J Exp Bot 44: Crane PR, EM Friis, KR Pedersen 1995 The origin and early diversification of angiosperms. Nature 374: Crane PR, S Lidgard 1989 Angiosperm diversification and paleolatitudinal gradients in Cretaceous floristic diversity. Science 246: Cronquist A 1988 The evolution and classification of flowering plants. New York Botanical Garden, Bronx. 555 pp. Donoghue MJ 1989 Phylogenies and the analysis of evolutionary sequences, with examples from seed plants. Evolution 43: Donoghue MJ, JA Doyle 1989 Phylogenetic analysis of angiosperms and the relationships of Hamamelidae. Pages in PR Crane, S Blackmore, eds. Evolution, systematics, and fossil history of the Hammamelidae. Vol 1. Clarendon, Oxford. Doyle JA 1978 Origin of angiosperms. Annu Rev Ecol Syst 9: Phylogeny of vascular plants. Annu Rev Ecol Syst 29: Doyle JA, MJ Donoghue 1986 Seed plant phylogeny and the origin of angiosperms: an experimental cladistic approach. Bot Rev 52: Phylogenies and angiosperm diversification. Paleobiology 19: Endress PK 1983 Dispersal and distribution in some small archaic relict angiosperm families (Austrobaileyaceae, Eupomatiaceae, Himantandraceae, Idiospermoideae-Calycanthaceae). Sonderbd Natwiss Ver Hambg 7: Enright NJ, D Goldblum 1998 Stand structure of the emergent conifer Agathis ovata in forest and maquis, Province Sud, New Caledonia. J Biogeogr 25: Ewers FW, JB Fisher, K Fichtner 1991 Water flux and xylem structure

8 712 INTERNATIONAL JOURNAL OF PLANT SCIENCES in vines. Pages in FE Putz, HA Mooney, eds. The biology of vines. Cambridge University Press, Cambridge. Feild TS, NM Holbrook In press Xylem sapflow and stem hydraulics of the vesselless angiosperm Drimys granadensis (Winteraceae) in a Costa Rican elfin forest. Plant Cell Environ. Feild TS, MA Zwieniecki, MJ Donoghue, NM Holbrook 1998 Stomatal plugs of Drimys winteri (Winteraceae) protect leaves from mist but not drought. Proc Natl Acad Sci USA 95: Frost FH 1930 Specialization in secondary xylem of angiosperms. I. Origin of vessel. Bot Gaz 89: Gartner BL 1991 Stem hydraulic properties of vines vs. shrubs of western poison oak, Toxicodendron diversilobum. Oecologia 87: Herendeen PS, EA Wheeler, P Baas 1999 Angiosperm wood evolution and the potential contribution of paleontological data. Bot Rev 65: Hickey LJ, JA Doyle 1977 Early Cretaceous fossil evidence for angiosperm evolution. Bot Rev 43: Jaffré T 1980 Etude ecologique du peuplement vegetal des sols derives de roches ultrabasiques en Nouvell Caledonie. Travaux de Documents 124. ORSTOM, Paris. 273 pp Distribution and ecology of the conifers of New Caledonia. Pages in NJ Enright, RS Hill, eds. Ecology of the southern conifers. Smithsonian Institution Press, Washington, D.C. Jérémie J 1980 Amborellaceae. Flore Nouvelle Caledonie Dependances 10: Mathews S, MJ Donoghue 1999 The root of angiosperm phylogeny inferred from duplicate phytochrome genes. Science 286: Metcalfe CR 1987 Anatomy of the dicotyledons. Vol 3. Magnoliales, Illiciales and Laurales. 2d ed. Clarendon, Oxford. METEO-France 1994 Climat de Nouvelle Caledonie. METEO-France en Nouvelle Caledonie, Noumeá. O Brien TP, ME McCully 1981 The study of plant structure: principles and selected methods. Termarcarphi, Melbourne. 357 pp. Parkinson CL, KL Adams, JD Palmer 1999 Multigene analyses identify the three lineages of extant flowering plants. Curr Biol 9: Pole M 1994 The New Zealand flora entirely long-distance dispersal? J Biogeogr 21: Qiu Y-L, J Lee, F Bernasconi-Quadroni, DE Soltis, PS Soltis, M Zanis, EA Zimmer, Z Chen, M Savolainen, MW Chase 1999 The earliest angiosperms: evidence from mitochondrial, plastid, and nuclear genomes. Nature 402: Sanderson MJ, MJ Donoghue 1994 Shifts in diversification rate with the origin of angiosperms. Science 264: Schneider EL, S Carlquist 1995a Vessels in the roots of Barclaya roundifolia (Nymphaeaceae). Am J Bot 82: b Vessel origins in Nymphaeaceae: Euryale and Victoria. Bot J Linn Soc 119: Smith AC 1947 The families Illiciaceae and Schisandraceae. Sargentia 7: Soltis PS, DE Solits, MW Chase 1999 Angiosperm phylogeny inferred from multiple genes as a tool for comparative biology. Nature 402: Sperry JS, JR Donnelly, MT Tyree 1988 A method for measuring hydraulic conductivity and embolism in xylem. Plant Cell Environ 11: Stebbins GL 1974 Flowering plants: evolution above the species level. Harvard University Press, Cambridge, Mass. 399 pp. Takahashi A 1988 Morphology and ontogeny of stem xylem elements in Sarcandra glabra (Thunb.) Nakai (Chloranthaceae): additional evidence for the occurrence of vessels. Bot Mag 101: Takhtajan AL 1969 Flowering plants: origin and dispersal. Smithsonian Institution Press, Washington, D.C. 310 pp. Taylor DW, LJ Hickey 1992 Phylogenetic evidence for the herbaceous origin of angiosperms. Plant Syst Evol 180: Thorne RF 1974 A phylogenetic classification of the Annoniflorae. Aliso 8: Tyree MT, FW Ewers 1991 The hydraulic architecture of trees and other woody plants. New Phytol 119: Hydraulic architecture of wood tropical plants. Pages in SS Mulkey, RL Chazdon, AP Smith, eds. Tropical forest ecophysiology. Chapman & Hall, New York. Tyree MT, B Sinclair, P Lu, A Granier 1993 Whole shoot hydraulic resistance in Quercus species measured with a new high-pressure flowmeter. Ann Sci For 50: Wing SL, LD Boucher 1998 Ecological aspects of the Cretaceous flowering plant radiation. Annu Rev Earth Planet Sci 26: Young DA 1981 Are the angiosperms primitively vesselless? Syst Bot 6: Zimmermann MH 1983 Xylem structure and the ascent of sap. Springer, Berlin. 143 pp. Zimmermann MH, AA Jeje 1981 Vessel-length distribution in stems of some American woody plants. Can J Bot 59: Zwieniecki MA, L Hutyra, MV Thompson, NM Holbrook 2000 Dynamic changes in petiole conductivity in red maple (Acer rubra L.), tulip tree (Liriodendron tulipifera L.), and northern fox grape (Vitis labrusca L.). Plant Cell Environ 23: